R_REDDYX.XYZ
AIC★ 51,088⑂ 6,095 forksApache-2.0

JustVugg/colibri

Tiered MoE Inference Engine

Runs large MoE models on commodity hardware by tiering experts: hot on GPU (CUDA/Vulkan), warm in RAM, cold streamed from NVMe.

// KEY FEATURES

  • Tiered expert storage across GPU, RAM, and NVMe
  • CUDA and Vulkan backends for GPU expert execution
  • Zero-dependency C implementation runs on existing hardware
Open on GitHub →

New repositories every 30 minutes

REDDYX AI scans GitHub 24/7 and ships the best AI/ML/Web3 projects to Telegram.

Join on Telegram
← Full catalog·Full index

// SIMILAR REPOSITORIES

HERDR - Agent Multiplexer in Terminal★ 51095Claude Code Practices★ 52343whisper.cpp: C/C++ port of Whisper★ 49828AGENT SKILL LIST★ 49756OBSIDIAN SKILLS AGENTS★ 52547LlamaIndex Data Indexing★ 49363

← FULL CATALOG

JustVugg/colibri — Tiered MoE Inference Engine | REDDYX AI